Allocating power between overhead, backup and computing power services

Abstract

The power controller allocates input power to the data center between the computing power for the computing service, the backup power, and the overhead power for the overhead system. The power controller reallocates overhead power and / or backup power to the computational power. This may increase the overall utilization of the data center by allowing additional processing power of the server to be used.

Description

Background Art

[0001] A data center typically includes a plurality of servers arranged in racks. A single data center may include a plurality of racks. A server may operate a virtual machine that may perform remote computing activities for one or more clients. A server consumes power based on the utilization of the server. The utilization may be based on the configuration of the virtual machine. For example, a virtual machine may be arranged as a computing node, which may utilize increased computing power and thus increase power consumption. In some cases, the utilization of a data center (e.g., the amount of computing capacity used by a data center) may be based on the power supply to the data center. For example, a processor may be limited by the amount of power supplied to it.

[0002] The data center may allocate input power to various services at the data center. For example, at least a portion of the input power may be allocated to provide power to a computing system. Some of the input power may be allocated to provide power to overhead systems of the data center, such as lighting, HVAC, cooling, etc. Some of the input power may be allocated to maintain standby power for a computing system. Summary of the invention

[0003] In some aspects, the technology described herein relates to a method for power allocation, including determining a power allocation for a computing system, the power allocation including a computing power allocation, a power redundancy allocation, and an overhead power allocation. The method includes determining a computing load on the computing system and reallocating at least a portion of the overhead power allocation based on the computing load.

[0004] In some aspects, the techniques described herein relate to a power distribution system for a data center, including one or more computing systems having computing capacity. The power distribution system includes input power, the input power is allocated between computing power allocations configured to provide power to the one or more computing systems, a first portion of the input power is allocated for computing power of the one or more computing systems, and a second portion of the computing power allocation is allocated for standby power and overhead power allocations. A power controller is configured to reallocate at least a portion of the overhead power allocation to the computing power allocation based on additional computing capacity of the one or more computing systems.

[0005] This Summary is provided to introduce a series of concepts that are further described in the Detailed Description below. This Summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used as an aid to limit the scope of the claimed subject matter.

[0006] Additional features and advantages of the embodiments of the present disclosure will be set forth in the following description, and some features and advantages will be apparent from the description, or may be learned by the practice of these embodiments. The features and advantages of these embodiments may be realized and obtained by the instruments and combinations particularly pointed out in the appended claims. These and other features will become more apparent from the following description and the appended claims, or may be learned by the practice of these embodiments set forth below. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to describe the manner in which the above and other features of the present disclosure can be obtained, a more specific description will be provided by referring to the specific implementations shown in the accompanying drawings. For better understanding, the same elements are designated by the same reference numerals in the various drawings. Although some of the drawings may be schematic or exaggerated representations of concepts, at least some of the drawings may be drawn to scale. It is to be understood that the drawings depict some example implementations, which will be described and explained with additional specificity and detail through the use of the drawings, in which:

[0008] Figure 1 is a representation of a power distribution system according to at least one embodiment of the present disclosure;

[0009] Figure 2 is a representation of a power diagram for a power distribution system according to at least one embodiment of the present disclosure;

[0010] Figure 3 is a representation of a power controller according to at least one embodiment of the present disclosure;

[0011] Figure 4-1 to Figure 4-5 A graph of power consumption at a data center according to at least one embodiment of the present disclosure is shown;

[0012] Figure 5 is a representation of an illustrative power distribution system for a data center in accordance with at least one embodiment of the present disclosure;

[0013] Figure 6 is a flowchart of a method for power allocation according to at least one embodiment of the present disclosure; and

[0014] Figure 7 is a representation of a computing system as a flow chart of a method for power allocation in accordance with at least one embodiment of the present disclosure. DETAILED DESCRIPTION

[0015] The present disclosure generally relates to devices, systems and methods for power management and allocation at data centers. Data centers use power to operate computing devices (such as racks, servers, virtual machines, storage devices, etc.) and to operate overhead equipment, such as lights, HVAC, computing cooling systems, etc. A power controller distributes input power to various elements of the data center. For example, a power controller can allocate part of the power of the backup power to be used as computing power. The power controller can allocate part of the power of the overhead power to other uses. For example, the power controller can allocate part of the power of the overhead power as computing power. In some examples, the power controller can allocate part of the energy storage (such as batteries in the backup power) back to the power grid. This can allow increased input power utilization.

[0016] According to at least one embodiment of the present disclosure, a power controller can allow a data center to utilize a higher proportion of the input power to the data center for computing resources. The computing system may have increased computing needs, which can utilize additional power. As discussed in further detail herein, in order to meet the increased computing needs, the power controller can reallocate at least part of the overhead power to the computing device. For example, the power controller can reallocate power from the HVAC system to meet the increased computing needs. In some examples, the power controller can reallocate power from the computing cooling system to meet the increased computing needs. This can increase the computing capacity of the data center.

[0017] In some embodiments, the power controller can reallocate portions of the overhead power for computational purposes. For example, the overhead power can include an allocation for recharging batteries for an uninterruptible power supply (UPS). When the UPS batteries are fully charged, the power allocation can be unutilized, and the power controller can reallocate the power for computational purposes. This can help reduce the amount of unused power at the data center.

[0018] In some embodiments, a power controller may receive input power source information and allocate power based on the input power source information. For example, a power controller may determine that the input power is primarily provided by a particular power source (e.g., renewable power, nuclear power, fossil fuel power), and the power controller may increase or decrease computing device utilization. This may allow a data center to operate flexibly based on a power source. In some examples, a power controller may determine a power price at a particular time and allocate power based on the power price. For example, if the power price is low, the utilization (and associated power consumption) of the data center may increase. If the power price is high, the power controller may sell the generated or stored power back to the power grid. This may increase the flexibility of the operation of the data center.

[0019] As shown by the foregoing discussion, the present disclosure utilizes various terms to describe the features and advantages of power distribution systems. Additional details about the meaning of these terms are now provided. For example, as used herein, the term "computing power" refers to the power utilized by a computing device for computing purposes. Specifically, the term computing power can include the power of a processor or server used to operate a data center. A server can operate a virtual machine in a computing mode, a storage mode, any other mode, and a combination thereof. In some embodiments, computing power can be used to supply power to a power source unit (PSU), an uninterruptible power supply (UPS), a backup power, a direct power source line to a server, any other server power, and a combination thereof.

[0020] The term "backup power" may refer to power used to provide power to a computing service in the event that at least a portion of the input power is disconnected from the computing service and / or server. In some embodiments, the backup power may be a separate power system, such as a backup power generator, a battery, or other backup power system.

[0021] In some embodiments, a portion of the input power may be allocated as power redundancy. Power redundancy may be a total power source reserved for a UPS or other input power system.

[0022] The term "overhead power" refers to power used for overhead systems or non-computing power services at a data center. For example, overhead power can refer to lighting, heating, ventilation, and air conditioning (HVAC), plumbing, management, computing cooling systems, UPS charging, power loss reserves, any other overhead power, and combinations thereof.

[0023] Figure 1 1 is a representation of an environment for a power distribution system 100 according to at least one embodiment of the present disclosure. The power distribution system 100 may be located at a data center 102, a colo (e.g., a colocation facility), or other system that includes one or more servers 104. In some embodiments, the data center may be a remote computing facility. Multiple servers 104 may be arranged in racks 106. The servers 104 may operate one or more virtual machines that may provide services to clients, such as computing services, storage services, etc.

[0024] The data center 102 may also include an overhead system 108. As discussed herein, the overhead system 108 may include services used to support the data center 102. For example, the overhead system 108 may include lighting, HVAC, computing cooling systems, battery charging, etc. In some embodiments, any loss associated with non-operating or non-currently prioritized equipment may include backup services. In some embodiments, the overhead system 108 may include any power consuming equipment that is not used for computing purposes. In some embodiments, power distribution redundancy may be included in the overhead system 108.

[0025] The power distribution system 100 may include an input power source 110. The input power source 110 may be input power to the data center 102. The input power source 110 may be from a power grid. For example, the input power source 110 may be supplied by a utility company or other power source. In some embodiments, the input power source 110 may be supplied on-site, such as by a generator, a field-based power plant, solar panels, etc. In some embodiments, the input power source 110 may include multiple input power lines based on different power sources and / or to provide increased reliability.

[0026] The power controller 112 may be used to allocate the input power source 110 to various services within the data center 102. The power controller 112 may allocate the input power from the input power source 110 to the servers 104 and / or the overhead systems 108. For example, the input power from the input power source 110 may have a total wattage or a total amount of wattage that may be used to provide power to the data center 102. The power controller 112 may allocate a first portion of the total wattage as a computing power 114 allocation to the servers 104 through one or more UPSs. As will be discussed in further detail herein, the power controller 112 may allocate a portion of the input power as a power redundancy allocation. In some embodiments, the power redundancy allocation may include a portion of the input power that is used to maintain computing power redundancy and overhead power requirements. For example, the power controller 112 may reserve a portion of the total power allocated for computing power 114 for power redundancy allocation, which may be utilized if another UPS loses functionality. The power controller 112 may allocate a second portion of the total wattage as overhead power 116 to the overhead systems 108. In this manner, power controller 112 may help provide power to servers 104 and / or overhead systems 108 for their respective services.

[0027] In accordance with at least one embodiment of the present disclosure, the power controller 112 may allocate power from the overhead power 116 to the compute power 114 allocation. In some cases, a portion of the total wattage allocated for the overhead power 116 may not be utilized. For example, weather conditions may allow a management facility and / or warehouse housing the racks 106 to operate without HVAC services and / or lighting. If the HVAC services and / or lighting are not used, a portion of the overhead power 116 allocated for these services may not be used. In order to reduce the amount of unused input power, the power controller 112 may reallocate a portion of the overhead power 116 allocated for the HVAC and / or lighting services to the compute power 114. This may allow the servers 104 to operate at a higher utilization rate and / or allow more servers 104 to be used.

[0028] In some embodiments, the power controller 112 may perform a trade-off to determine the power allocation between the overhead power 116 and the compute power 114. For example, a client request may increase the workload on the server 104. The power controller 112 may determine that one or more of the overhead systems 108 may be temporarily suspended. The power controller 112 may reallocate power from the suspended services to the compute power 114 allocation. When the increased workload has completed or decreased, or when the suspended services continue, the power controller 112 may reallocate power back to the overhead power 116.

[0029] As an example, a computing cooling system may be used to maintain the server 104 at an operating temperature. Without the computing cooling system, the temperature of the server 104 may not increase immediately, but the increase in temperature may be gradual. In addition, the computing cooling system may include a tank or other reservoir of cooled cooling fluid. During periods of increased workload on the server 104, the computing cooling system may be suspended, and power for the computing cooling system may be reallocated to the computing power 114 allocation. When one or more of the increased workloads have been completed, the temperature of the server 104 has increased above a threshold, and the computing cooling system has run out of cooled cooling fluid, the power controller 112 may reallocate computing cooling system power from the computing power 114 back to the overhead power 116.

[0030] In some embodiments, the power controller 112 may receive a request for more computing capacity, and therefore, more computing power 114. For example, the data center 102 may receive an increased demand for computing services. The data center 102 may request additional computing capacity and / or computing power 114 from the power controller 112. Based on the request, the power controller 112 may reallocate at least a portion of the overhead power 116 to the computing power 114 to satisfy the request for additional power.

[0031] In some embodiments, the power controller 112 can allocate power based on the input power source 110. For example, some clients may desire to use a certain type of input power source 110 to supply their computing power 114, such as renewable power (e.g., solar, wind, hydroelectric, geothermal), fossil fuel power (e.g., coal, natural gas), nuclear power, any other type of input power source 110, and combinations thereof. In some embodiments, the power controller 112 can allocate power based on power prices. For example, power prices can fluctuate throughout the day and / or across different dates. When power prices are low, the power controller 112 can allocate power to charge batteries or perform other tasks. In some embodiments, when power prices are higher than power generation prices, the power controller 112 can utilize secondary power sources, such as on-site generators. In some embodiments, the power controller 112 can send power back to the power grid to profit from high grid power prices. In some embodiments, the power controller 112 can make power allocation choices based on the carbon or water impact of a particular power source. For example, the power controller 112 can select a low carbon impact power source for battery charging. In some embodiments, the power controller 112 may make selections for inputs based on a balance of inputs including energy prices, carbon shock, water shock, weather, computing spot markets, external factors (grid energy storage), any other inputs, and combinations thereof.

[0032] Figure 2 According to at least one embodiment of the present disclosure, a power distribution system (eg, Figure 1 2. The power distribution system 100 of FIG. 1 is a representation of a power graph 218 of a power distribution system 100 of FIG. 1. The power graph 218 may include one or more input power sources 210. The one or more input power sources 210 may provide power to a power controller 212. The power controller 212 may allocate power to computing power 214 and overhead power 216 for operation of the data center.

[0033] In some embodiments, the input power source 210 may include grid power 220. Grid power 220 may include any type of power source that is used to provide general grid power. For example, grid power 220 may include renewable power 222, nuclear power 224, fossil fuel power 226, any other grid power source, and combinations thereof. In some embodiments, one or more input power sources 210 may also include power on-site power 228. Power on-site power 228 may include any type of power on-site power, such as a diesel or gas powered generator, a solar panel, a power storage device (e.g., a battery), any other type of power on-site power 228, and combinations thereof.

[0034] The power controller 212 can allocate power from the grid power 220 and / or the power farm power 228 to the operations of the data center, including allocating power to one or more of the computing power 214, the backup power 232, and the overhead power 216. In some embodiments, as discussed herein, the power controller 212 can power allocate power from the overhead power 216 to the computing power 214 based on the utilization of the overhead power 216. In some embodiments, the power controller 212 can power allocate power from the power farm power 228 to the computing power 214 and / or the overhead power 216. In some embodiments, the power controller 212 can power allocate power from the power farm power 228 to the computing power 214 and / or the overhead power 216 based on one or more of the supplies from the renewable power 222, the nuclear power 224, and the fossil fuel power 226.

[0035] As discussed herein, in some embodiments, the power controller 212 may reallocate at least a portion of the overhead power 216 back to the grid power 220, or the power controller 212 may reallocate at least a portion of the overhead power back to the power grid via energy stored in a backup power device. For example, the power controller 212 may sell the stored power allocated to the overhead power 216 back to the grid. In some embodiments, the power controller 212 may reallocate at least a portion of the overhead power 216 back to the grid power 220. For example, the power controller 212 may sell the power allocated to the overhead power 216 back to the grid. In some embodiments, the power controller 212 may reallocate at least a portion of the power of the power farm power 228 to the grid power 220. For example, the power controller 212 may sell the power generated on-farm to the grid.

[0036] The power controller 212 may utilize any input to determine the power allocation between one or more of the overhead power 216, the compute power 214, the power farm power 228, and the grid power 220. For example, the power controller 212 may maximize the utilization of the computing resources of the data center. A data center may have a variable computing load based on client demand, consumer patterns, number of clients, etc. The power controller 212 may analyze the computing load and allocate power to the compute power 214 to increase the capacity of the data center.

[0037] In some examples, the power controller 212 can maximize the profitability of the data center by monitoring the operating costs of the data center, the cost of power from the grid power 220, the cost of power generation from the power farm 228, the utilization of computing equipment, the utilization of overhead systems powered by the overhead power 216, any other cost information, and combinations thereof. Monitoring cost information can help allocate power throughout the power graph 218. In some embodiments, the power controller 212 can make power allocation choices based on the carbon or water impact of a particular power source. For example, the power controller 212 can select a low carbon impact power source for battery charging. In some embodiments, the power controller 212 can make choices for inputs based on a balance of inputs, including energy prices, carbon impact, water impact, weather, computing spot markets, external factors (grid energy storage), any other inputs, and combinations thereof.

[0038] In some embodiments, the power controller 212 may receive a request for more computing capacity, and therefore, for more computing power 214. For example, a data center may receive an increased demand for computing services. The data center may request additional computing capacity and / or computing power 214 from the power controller 212. Based on the request, the power controller 212 may reallocate at least a portion of the overhead power 216 to the computing power 214 (such as by suspending one or more overhead systems) to satisfy the request for additional power.

[0039] Figure 3 is a representation of a power controller 312 according to at least one embodiment of the present disclosure. Each of the components of the power controller 312 may include software, hardware, or both. For example, a component may include one or more instructions stored on a computer-readable storage medium and executable by a processor of one or more computing devices (such as a client device or a server device). When executed by one or more processors, the computer-executable instructions of the power controller 312 may cause (multiple) computing devices to perform the methods described herein. Alternatively, the component may include hardware, such as a dedicated processing device that performs a specific function or group of functions. Alternatively, the components of the power controller 312 may include a combination of computer-executable instructions and hardware.

[0040] In addition, for example, the components of the power controller 312 can be implemented as one or more operating systems, one or more independent applications, one or more modules of an application, one or more plug-ins, one or more library functions or functions that can be called by other applications, a cloud computing model, or a combination thereof. Therefore, these components can be implemented as independent applications, such as desktop or mobile applications. In addition, these components can be implemented as one or more web-based applications hosted on a remote server. These components can also be implemented in a set of mobile device applications or "apps".

[0041] The power controller 312 includes a power allocator 330. The power allocator 330 can allocate input power into the data center between services at the data center. For example, the power allocator 330 can allocate power between a computing power 314 allocation, a backup power 332 allocation, and an overhead power 316 allocation. The power controller 312 includes a computing power analyzer 334. The computing power analyzer 334 can analyze and / or predict the utilization of servers and other computing systems of the data center. Using the utilization, the computing power analyzer 334 can determine the power consumption of the servers and computing devices.

[0042] The power controller 312 includes an overhead power analyzer 336. The overhead power analyzer 336 can analyze the power utilization of one or more overhead systems powered by the overhead power 316. For example, the overhead power analyzer 336 can analyze the power utilization of the cooling system 338 of the data center. The cooling system 338 can be any type of cooling system 338. For example, the cooling system 338 can be a fluid-cooled cooling system. The cooling system 338 can include a reservoir or tank of a cooled cooling fluid (e.g., a cooling fluid at a temperature for effectively cooling a computing device). The overhead power analyzer 336 can determine the amount of cooling that the cooled cooling fluid can perform. Based on the reservoir of the cooled cooling fluid, the power allocator 330 can allocate the power used by the cooling system 338 to the computing power 314. In this way, when the cooled cooling fluid cools the computing device, the computing capacity of the data center can be increased. In this way, the cooling system 338 can use the thermal energy stored in the cooling fluid reservoir to increase the computing capacity of the data center.

[0043] In some embodiments, cooling system 338 can be any other type of cooling system. For example, cooling system 338 can be air cooling or a fan-based cooling system. The fans can be turned off to temporarily increase the computing capacity of the data center. This can increase the operating temperature of the computing devices for a period of time. When power distributor 330 can distribute power back to cooling system 338 powered by overhead power 316, the computing capacity can be increased until the operating temperature of the computing devices rises above the threshold temperature.

[0044] The overhead system may include one or more batteries 340. For example, the battery may be connected to a UPS, a secondary power source, a primary power source, any other battery, or a combination thereof, or may be part of a UPS, a secondary power source, a primary power source, any other battery, or a combination thereof. The battery 340 may be charged using the overhead power 316, and the power distributor 330 may allocate input power to the overhead power 316 for the charging. Once the battery 340 is charged, the portion of the overhead power 316 may not be used. The overhead power analyzer 336 may determine the amount of power used to charge the battery 340. If the battery is fully charged, the overhead power analyzer 336 may determine that the portion of the overhead power 316 used to charge the battery 340 may not be used. The overhead power analyzer 336 may communicate the utilization of the battery charging power to the power distributor 330. The power distributor 330 may allocate the portion of the overhead power 316 to the computing power 314, thereby increasing the computing capacity of the data center. In some embodiments, the power distributor 330 may allocate the power stored in the battery 340 to the computing power 314.

[0045] The overhead power analyzer 336 may also analyze the power used by the lighting and HVAC 342 of the management office and other auxiliary support facilities. Based on the needs calculated by the data center, the power allocator 330 may determine that the power allocated by the lighting and HVAC 342 may be allocated to the computing power 314. In some embodiments, the overhead power analyzer 336 may use external information to determine whether the lighting and HVAC 342 power can be reallocated to the computing power 314. For example, when the management staff is not at the data center, the overhead power analyzer 336 may determine that the lighting and HVAC 342 may not be used on weekends or holidays. In some examples, the overhead power analyzer 336 may determine the weather so that the lighting and HVAC 342 may not be used to maintain a comfortable working environment. In some embodiments, the overhead power analyzer 336 may analyze the power used by any other power system, including backup computing systems, network systems, redundant network equipment, any other power system, and combinations thereof.

[0046] In some embodiments, the overhead power analyzer 336 can analyze the amount of overhead power 316 allocated for power losses 344 in the system. The power allocator 330 can allocate a portion of the input power to the power losses 344. These power losses 344 can be losses in the power and power systems of the data center. The power losses 344 can be based on the total power usage of the data center. In other words, as power usage increases, the power losses 344 may increase. As power usage decreases, the power losses 344 may decrease. In some embodiments, based on the amount of total power used by the data center, the power allocator 330 can reallocate a portion of the input power allocated to the power losses 344 to the computing power 314 or another power system.

[0047] The power controller 312 may include a power source analyzer 346. The power allocator 330 may allocate power between the computing power 314, the standby power 332, and the overhead power 316 based on information provided by the power source analyzer 346. For example, the power source analyzer 346 may determine that the input power is supplied by a specific power source, or that the input power is provided at a specific price, and allocate the input power based on the input power source. In some embodiments, the power allocator 330 may allocate more power to the overhead power 316 based on the power source or price. For example, thermal energy may be stored in a cooling system 338 and / or a battery 340 that is charged based on a specific power source and / or power price.

[0048] In some embodiments, computing power analyzer 334 may identify a request for additional computing capacity and / or additional computing power 314. Based on the request for additional computing capacity, power allocator 330 may review overhead power allocation based on overhead power analyzer 336. Power allocator 330 may determine the amount of additional computing capacity and / or additional computing power 314 requested. Based on the amount of additional computing capacity and / or additional computing power 314 requested, power allocator 330 may allocate power from one or more of cooling system 338, battery 340, lighting and HVAC 342, or power loss 334 to computing power 314. For example, power allocator 330 may suspend one or more of these overhead systems, such as cooling system 338, battery 340, lighting and HVAC 342, or power loss 344. This may help increase computing capacity based on the requested additional capacity.

[0049] Figure 4-14 is a power chart 448 that plots power on a vertical axis (e.g., y-axis) and time on a horizontal axis (e.g., x-axis) in accordance with at least one embodiment of the present disclosure. The power chart 448 includes multiple sections indicating power allocations, including a computing power 414 allocation, an overhead power 416 allocation, and a power redundancy 432 allocation. The power allocation may be power available for use by a computing system, or may be power that may be allocated for use by computing services at a data center.

[0050] A power consumption graph 450 is plotted over time on the power chart 448. When the power consumption graph 450 is in the overhead power 416, the power allocated to the overhead power 416 may be sufficient to power the computing system. As the power consumption of the computing system increases, the power consumption graph 450 may exceed the overhead power 416 allocation. In accordance with at least one embodiment of the present disclosure, the power controller may allocate power from one or both of the computing power 414 allocation or the power redundancy 432 allocation.

[0051] For example, in Figure 4-1 In the illustrated embodiment, as power consumption increases beyond the overhead power 416 allocation, as indicated by the power consumption graph 450 extending outside of the overhead power 416 allocation region, the power controller may allocate a portion of the power redundancy 432 allocation to meet the increased power consumption.

[0052] In some embodiments, the power consumption of the computing system may exceed the combined overhead power 416 allocation and the computing power 414 allocation, as can be seen from the power consumption graph 450. When the power consumption exceeds the combined overhead power 416 allocation and the computing power 414 allocation, the power controller may allocate power redundancy 432 to the computing power 414.

[0053] Figure 4-2 4 is a representation of a power graph 448 in which at least a portion of the overhead power 416 is unavailable for a period of time. The power graph 448 shows a maximum power capacity 452, which may be a combined power of the overhead power 416 allocation, the computing power 414 allocation, and the power redundancy 432 allocation. As can be seen, when a portion of the input power becomes unavailable for use by the computing system, the maximum power capacity 452 may be reduced. Because the power redundancy 432 allocation is still available, the maximum power capacity 452 is not reduced all the way to the computing power 414, but only the amount of the unavailable overhead power 416.

[0054] As discussed herein, the power controller may allocate overhead power 416 as unavailable for use by the computing system based on any factor, including time of day, a particular date, overhead system availability, power source, any other factors, and combinations thereof.

[0055] exist Figure 4-3 In the example, power redundancy 432 is used for unavailable power for use by the computing system for a period of time. For example, power redundancy 432 can be used as backup power (e.g., to support one or more UPSs that have lost power), and power redundancy 432 can be allocated to be unavailable for use by the computing system. Overhead power 416 is still available, and maximum power capacity 452 can be reduced based on the unavailability of power redundancy 432.

[0056] exist Figure 4-4 In the embodiment of the present invention, the overhead power 416 allocation available for use by the computing system can be variable. As discussed herein, the overhead power 416 allocation can be based on multiple overhead systems. The power controller can determine that some of the overhead systems should be used at a particular time, while other overhead systems may not be used, based on the requirements of the computing system.

[0057] As can be seen, the power redundancy 432 can have a constant allocation. In other words, the power redundancy 432 allocation can be all available or all unavailable. In this way, as can be seen, the maximum power capacity 452 can vary based on the allocation of the overhead power 416.

[0058] exist Figure 4-5 4, a power consumption graph 450 shows the power consumption of computing services of a data center. As can be seen, the power controller can allocate computing power 414, overhead power 416, and power redundancy 432 to power computing services. In some embodiments, power redundancy 432 may become unavailable for allocation to computing services over a period of time. As can be seen from power consumption graph 450, the computing system can utilize power redundancy 432.

[0059] When power redundancy 432 becomes unavailable, the power controller may perform one or more mitigation controls to reduce the power consumption of the computing system to below the maximum power capacity 452. For example, the power controller may adjust the computing service, cut off power to one or more servers, or otherwise reduce the power consumption of the computing service. This may be followed by calibration or other periods of time to ensure that the power consumption of the computing system remains below the maximum power capacity 452.

[0060] Figure 55 is a representation of an illustrative power distribution system 554 for a data center 502 in accordance with at least one embodiment of the present disclosure. A power controller 512 can receive an analysis input to determine power usage of the data center. The power controller 512 can provide the power usage analysis to a power allocator 530. The power allocator 530 can allocate power between computing power 514 and overhead power 516. In some embodiments, the power allocator 530 can allocate power to charge a UPS and / or other batteries that are part of the data center 502.

[0061] The power controller 512 can receive telemetry 556 regarding power consumption at the data center 502. For example, the telemetry 556 can include measurements from input power sensors 558, which can measure input power to the data center 502. The telemetry 556 can include measurements from compute power sensors 560, which can measure input power to the servers and racks of the data center 502. The telemetry 556 can include measurements from overhead power sensors 562, which can measure input power to overhead systems. The telemetry 556 can include measurements from rechargeable elements of the data center 502, including a UPS and / or any rechargeable batteries.

[0062] Using telemetry 556, power controller 512 can analyze the power consumption of computing systems and overhead systems. In some embodiments, power controller 512 can also include workload data 564. Workload data 564 can include information about the current computing load of the computing system. For example, workload data 564 can include information about processing operations performed by servers and computing devices. This can help further understand the current power consumption of computing power 514 and / or identify trends in power consumption.

[0063] In some embodiments, the power controller can receive prediction information for future power consumption from the power predictor 566. The power predictor 566 can predict the amount of computing power that can be used by the computing center and / or the amount of overhead power used by the overhead system. For example, the power predictor 566 can determine the predicted computing power based on the order of processing resources from one or more customers, based on historical trends, based on current trends, any other mechanism and combinations thereof. In some examples, the power predictor 566 can determine the predicted overhead power based on any factor (such as weather, predicted computing load, any other factor and combinations thereof). In some embodiments, the power predictor 566 can utilize one or more external factors 568 (such as weather, power source, etc.) to predict computing power and / or overhead power. The power controller 512 can include the predicted power determination in its power analysis, and the power allocator 530 can allocate power based on the predicted power.

[0064] In some embodiments, the power controller 512 may incorporate other information into the power analysis. For example, the power controller 512 may include an inventory 570 of computing devices. The inventory 570 may include a physical inventory of the number of servers. The inventory 570 may include an inventory of the processing capacity of each server. The inventory 570 may include an inventory of the total processing capacity of the data center 502. In some embodiments, the inventory 570 may maintain an inventory of the number of servers that are inoperable or operating at less than full capacity. This may allow the power controller 512 to determine the total available processing capacity of the data center 502.

[0065] In some embodiments, the power controller 512 can be at least partially affected by the power level 572. As will be discussed in further detail herein, the power level 572 can include the level of power sources that will be used based on certain conditions. For example, the power level 572 can include instructions for which local power sources to use if the main power source is powered off, including instructions for how long to use each power source after the power outage event. Different power sources can be set for different situations. For example, in order to maintain operation within the first minute of the main power source being powered off, a supercapacitor or other fast discharge power source can be used. For short periods of time, such as between the first minute and the first hour after a power outage, a battery or other power storage element can be used. For one hour to 48 hours (or longer), a diesel or other fossil fuel-powered generator can be used.

[0066] In some embodiments, the power controller 512 can adjust the power level 572 based on any factors. For example, the power controller 512 can adjust the power level 572 based on the cost of input power, the cost of power within the power farm, the capital cost of the power source within the farm, etc. This can help maintain power continuity while reducing costs. In some embodiments, the power controller 512 can adjust the power level 572 based on the power allocation and / or utilization of the computing system.

[0067] Figure 6 , corresponding text, and examples provide several different methods, systems, devices, and non-transitory computer-readable media for power controllers and power distribution systems. In addition to the above, Figure 6 As shown in FIG. 1 , one or more embodiments may also be described in terms of flowcharts including acts for achieving certain results. Figure 6 More or fewer actions may be utilized to perform. In addition, these actions may be performed in different orders. Additionally, the actions described herein may be repeated or performed in parallel with each other, or repeated or performed in parallel with different instances of the same or similar actions.

[0068] As mentioned above, Figure 6 A flowchart of a method 674 or series of actions for managing power allocation at a data center is shown in accordance with one or more embodiments. Figure 6 Actions according to one embodiment are shown, but alternative embodiments may omit, add, reorder, and modify Figure 6 One or more of any actions shown in . Figure 6 The actions of may be performed as part of a method. Alternatively, a non-transitory computer readable medium may include instructions that, when executed by one or more processors, cause a computing device to perform Figure 6 In some embodiments, the system may perform Figure 6 action.

[0069] At 676, the power controller may identify a power allocation for the computing system. For example, the power controller may identify an allocation of power between a computing power allocation, a power redundancy allocation, an overhead power allocation, or a combination thereof. At 678, the power controller may determine a computing load on the computing system. For example, the power controller may determine whether the computing load on the computing system is less than, meets, or exceeds the available computing power based on the computing power allocation.

[0070] Based on the determined computational load, at 680, the power controller may reallocate at least a portion of the overhead power allocation. For example, the power controller may reallocate at least a portion of the overhead power allocation to the computational power allocation, such as when the computational load exceeds the computational power on the computing system. In some examples, the power controller may reallocate at least a portion of the overhead power allocation back to the grid. In this manner, the overhead power allocation may not be unused, thereby improving power efficiency of the data center.

[0071] Figure 7 Certain components are shown that may be included within computer system 700. One or more computer systems 700 may be used to implement the various devices, components, and systems described herein.

[0072] Computer system 700 includes processor 701. Processor 701 may be a general purpose single-chip or multi-chip microprocessor (e.g., Advanced RISC (Reduced Instruction Set Computer) Machine (ARM)), a dedicated microprocessor (e.g., a digital signal processor (DSP)), a microcontroller, a programmable gate array, etc. Processor 701 may be referred to as a central processing unit (CPU). Figure 7 Only a single processor 701 is shown in computer system 700, but in an alternative configuration, a combination of processors (e.g., an ARM and DSP) could be used.

[0073] The computer system 700 also includes a memory 703 in electronic communication with the processor 701. The memory 703 may be any electronic component capable of storing electronic information. For example, the memory 703 may be embodied as a random access memory (RAM), a read-only memory (ROM), a magnetic disk storage medium, an optical storage medium, a flash memory device in a RAM, on-board memory included in a processor, an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) memory, a register, etc. (including combinations thereof).

[0074] Instructions 705 and data 707 may be stored in memory 703. Instructions 705 may be executed by processor 701 to implement some or all of the functions disclosed herein. Executing instructions 705 may involve using data 707 stored in memory 703. Any of the various examples of modules and components described herein may be implemented in part or in whole as instructions 705 stored in memory 703 and executed by processor 701. Any of the various examples of data described herein may be among the data 707 stored in memory 703 and used by processor 701 during execution of instructions 705.

[0075] The computer system 700 may also include one or more communication interfaces 709 for communicating with other electronic devices. The communication interface(s) 709 may be based on wired communication technology, wireless communication technology, or both. Some examples of the communication interface 709 include a universal serial bus (USB), an Ethernet adapter, a wireless adapter operating according to the Institute of Electrical and Electronics Engineers (IEEE) 802.11 wireless communication protocol, A wireless communication adapter, and an infrared (IR) communication port.

[0076] The computer system 700 may also include one or more input devices 711 and one or more output devices 713. Some examples of input devices 711 include keyboards, mice, microphones, remote control devices, buttons, joysticks, trackballs, touchpads, and light pens. Some examples of output devices 713 include speakers and printers. One specific type of output device that is typically included in the computer system 700 is a display device 715. The display device 715 used with the embodiments disclosed herein can utilize any suitable image projection technology, such as a liquid crystal display (LCD), a light emitting diode (LED), gas plasma, electroluminescence, etc. A display controller 717 may also be provided for converting data 707 stored in the memory 703 into text, graphics, moving images, or a combination thereof (as appropriate) displayed on the display device 715.

[0077] The various components of computer system 700 may be coupled together by one or more buses, which may include a power bus, a control signal bus, a status signal bus, a data bus, etc. For clarity, the Figure 7 Various buses are shown as bus system 719 .

[0078] One or more specific embodiments of the present disclosure are described herein. These described embodiments are examples of the currently disclosed technology. Additionally, in order to provide a concise description of these embodiments, not all features of the actual embodiments are described in the specification. It should be understood that in the development of any such actual implementation, as in any engineering or design project, many implementation-specific decisions will be made to achieve the specific goals of the developers, such as complying with system-related and business-related constraints, which may vary from embodiment to embodiment. In addition, it should be understood that such development work can be complex and time-consuming, but for those of ordinary skill in the art who benefit from the present disclosure, it will be a routine task of design, manufacture, and production.

[0079] The articles "one", "an" and "the" are intended to mean that there are one or more elements in the elements in the foregoing description. The terms "comprising", "including", and "having" are intended to be inclusive, and mean that there may be additional elements in addition to the listed elements. Additionally, it should be understood that reference to "one embodiment" or "embodiment" of the present disclosure is not intended to be interpreted as excluding the existence of additional embodiments that also include the listed features. For example, any element described in relation to the embodiments herein can be combined with any element of any other embodiment described herein. As will be understood by a technician of ordinary skill in the art covered by the embodiments of the present disclosure, the numerals, percentages, ratios or other values ​​specified herein are intended to include the value, as well as other values ​​that are "approximately" or "approximately" to the specified value. Therefore, the specified value should be interpreted broadly enough to cover values ​​that are at least close enough to the specified value to perform the desired function or achieve the desired result. The stated values ​​at least include expected variations from appropriate manufacturing or production processes, and may include values ​​within 5%, within 1%, within 0.1%, or within 0.01% of the stated value.

[0080] Those of ordinary skill in the art will recognize that various changes, substitutions, and modifications may be made to the embodiments disclosed herein in view of the equivalent constructions of the present disclosure without departing from the spirit and scope of the present disclosure, and without departing from the spirit and scope of the present disclosure. The equivalent construction of the "means plus function" clause including functions is intended to cover structures described herein as performing the functions listed, including structural equivalence that operates in the same manner, and equivalent structures that provide the same functions. Except for those claims in which the words "method for..." appear together with the associated functions, it is the express intention of the applicant not to invoke means plus function or other functional claims for any claim. Each addition, deletion, and modification to the embodiments that falls within the meaning and scope of the claims will be encompassed by the claims.

[0081] As used herein, the terms "approximately," "about," and "substantially" refer to an amount that is close to a specified amount and still performs a desired function or achieves a desired result. For example, the terms "approximately," "about," and "substantially" may refer to an amount that is within less than 5%, less than 1%, less than 0.1%, and less than 0.01% of a specified amount. In addition, it should be understood that any directions or reference frames in the foregoing description are only relative directions or movements. For example, any reference to "upward" and "downward" or "above" or "below" is only a description of the relative position or movement of the associated elements.

[0082] Without departing from its spirit or characteristics, the present disclosure may be embodied in other specific forms. The described embodiments are to be considered illustrative rather than restrictive. Therefore, the scope of the present disclosure is indicated by the appended claims rather than by the above description. Changes within the equivalent meaning and scope of the claims should be included within their scope.

Claims

1. A method for power distribution, comprising: Determining a power allocation for a computing system, the power allocation comprising a computing power allocation, a power redundancy allocation, and an overhead power allocation; determining a computational load on the computing system; as well as At least a portion of the overhead power allocation is reallocated based on the computational load. 2 . The method of claim 1 , wherein reallocating the portion of the overhead power allocation comprises reallocating the portion of the overhead power allocation to the compute power allocation. 3 . The method of claim 1 , wherein reallocating the portion of the overhead power allocation comprises reallocating the portion of the overhead power allocation to a power grid.

4. The method of claim 1, further comprising receiving a request for additional computing capacity, and wherein reallocating the portion of the overhead power allocation comprises reallocating the portion of the overhead power allocation to the computing power allocation.

5. The method of claim 1 further comprising receiving a request for additional computing capacity, and wherein reallocating the portion of the overhead power allocation comprises reallocating at least a portion of the power redundancy allocation and the portion of the overhead power allocation to the computing power allocation.

6. The method of claim 1, wherein reallocating the portion of the overhead power allocation comprises suspending power to overhead systems.

7. The method of claim 6, wherein suspending power to the overhead system comprises suspending power to a computing cooling system.

8. The method of claim 6, wherein suspending power to the overhead systems comprises suspending power to lighting and HVAC.

9. A system for power distribution, comprising: Memory; as well as A processor comprising instructions causing the processor to: Determining a power allocation for a computing system, the power allocation comprising a computing power allocation, a power redundancy allocation, and an overhead power allocation; determining a computational load on the computing system; and At least a portion of the overhead power allocation is reallocated based on the computational load.

10. The system of claim 9, wherein reallocating the portion of the overhead power allocation comprises reallocating the portion of the overhead power allocation to the compute power allocation.

11. The system of claim 9, wherein reallocating the portion of the overhead power allocation comprises reallocating a portion of the stored reserve power to a power grid.

12. The system of claim 9, wherein the instructions further cause the processor to receive a request for additional computing capacity, and wherein reallocating the portion of the overhead power allocation comprises reallocating the portion of the overhead power allocation to the computing power allocation.

13. The system of claim 9, wherein the instructions further cause the processor to receive a request for additional computing capacity, and wherein reallocating the portion of the overhead power allocation comprises reallocating at least a portion of the backup power allocation and the portion of the overhead power allocation to the computing power allocation.

14. The system of claim 9, wherein the instructions further cause the processor to determine a predicted computing power, and wherein determining the power allocation comprises determining the power allocation based on the predicted computing power.

15. The system of claim 14, wherein determining the predicted computing power comprises determining the predicted computing power based on one or more of a sequence, a historical trend, or a current trend of processing resources.

Images